Bayesian graph convolutional neural networks for semi-supervised classification

Zhang, Yingxue; Pal, Soumyasundar; Coates, Mark; Üstebay, Deniz

doi:10.48550/arxiv.1811.11103

Cited by 6 publications

(7 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Most methods from the above two classes treat the graph as fixed observation and only a few methods [16,22,12] treat the graph as a random variable and model it with generative models. Among them, Ng et al [16] focus more on an active learning setting on graphs; Zhang et al [22] models the graph along with a stochastic block model and does not consider the interaction between the graph and the features or the labels in the generative model of the graph; the work of Liu [12] is the most similar one as our framework, but is restricted in the linear models. To our best knowledge, our work is the first to propose a generative framework for graph-based semi-supervised learning that models the joint distribution of features, outcomes, and the graph with flexible nonlinear models.…”

Section: Generative Methods For Graph-based Semi-supervised Learningmentioning

confidence: 99%

“…A few previous work [22,12] trying to apply generative models in graph-based semi-supervised learning have been restricted to relatively simple model families due to the difficulty in efficient training of generative models. Thanks to recent advances of scalable variational inference techniques [8,7], we are able to propose a flexible generative framework for graph-based semisupervised learning.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Flexible Generative Framework for Graph-based Semi-supervised Learning

Ma,

Tang,

Zhu

et al. 2019

Preprint

View full text Add to dashboard Cite

We consider a family of problems that are concerned about making predictions for the majority of unlabeled, graph-structured data samples based on a small proportion of labeled examples. Relational information among the data samples, often encoded in the graph or network structure, is shown to be helpful for these semisupervised learning tasks. Conventional graph-based regularization methods and recent graph neural networks do not fully leverage the interrelations between the features, the graph, and the labels. We propose a flexible generative framework for graph-based semi-supervised learning, which approaches the joint distribution of the node features, labels, and the graph structure. Borrowing insights from random graph models in network science literature, this joint distribution can be instantiated using various distribution families. For the inference of missing labels, we exploit recent advances of scalable variational inference techniques to approximate the Bayesian posterior. We conduct thorough experiments on benchmark datasets for graph-based semi-supervised learning. Results show that the proposed methods outperform state-of-the-art models under most settings.

show abstract

Section: Generative Methods For Graph-based Semi-supervised Learningmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A Flexible Generative Framework for Graph-based Semi-supervised Learning

Ma,

Tang,

Zhu

et al. 2019

Preprint

View full text Add to dashboard Cite

show abstract

“…We note that recent work [33] also claims to estimate uncertainty in the graph convolutional neural networks setting. This work uses a two-stage strategy: it firstly takes the observed network as a realisation from a parametric Bayesian relational model, and then uses Bayesian Neural Networks to infer the model parameters.…”

Section: Related Workmentioning

confidence: 99%

Scalable Deep Generative Relational Models with High-Order Node Dependence

Fan,

Li,

Sisson

et al. 2019

Preprint

View full text Add to dashboard Cite

We propose a probabilistic framework for modelling and exploring the latent structure of relational data. Given feature information for the nodes in a network, the scalable deep generative relational model (SDREM) builds a deep network architecture that can approximate potential nonlinear mappings between nodes' feature information and the nodes' latent representations. Our contribution is two-fold:(1) We incorporate high-order neighbourhood structure information to generate the latent representations at each node, which vary smoothly over the network. (2) Due to the Dirichlet random variable structure of the latent representations, we introduce a novel data augmentation trick which permits efficient Gibbs sampling. The SDREM can be used for large sparse networks as its computational cost scales with the number of positive links. We demonstrate its competitive performance through improved link prediction performance on a range of real-world datasets.

show abstract

“…This function is often approximated using a fixed neural network that provides a likely correlational explanation for the relationship between x and y. There has been good work done in this area [12] and there are two main types of Bayesian Deep Learning. This is not an exhaustive list of all the methods, but a broad overview of two.…”

Section: Bayesian Deep Learningmentioning

confidence: 99%

“…This integral is often intractable in practice, and number of techniques have been proposed in the literature to overcome [12] this problem. Intuitively, a Bayesian neural network is encoding information about the distribution over output values given certain inputs.…”

Section: Bayesian Deep Learningmentioning

confidence: 99%

Graph Neural Processes: Towards Bayesian Graph Neural Networks

Carr¹,

Wingate²

2019

Preprint

View full text Add to dashboard Cite

We introduce Graph Neural Processes (GNP), inspired by the recent work in conditional and latent neural processes. A Graph Neural Process is defined as a Conditional Neural Process that operates on arbitrary graph data. It takes features of sparsely observed context points as input, and outputs a distribution over target points. We demonstrate graph neural processes in edge imputation and discuss benefits and draw backs of the method for other application areas. One major benefit of GNPs is the ability to quantify uncertainty in deep learning on graph structures. An additional benefit of this method is the ability to extend graph neural networks to inputs of dynamic sized graphs.Preprint. Under review.

show abstract

Bayesian graph convolutional neural networks for semi-supervised classification

Cited by 6 publications

References 0 publications

A Flexible Generative Framework for Graph-based Semi-supervised Learning

A Flexible Generative Framework for Graph-based Semi-supervised Learning

Scalable Deep Generative Relational Models with High-Order Node Dependence

Graph Neural Processes: Towards Bayesian Graph Neural Networks

Contact Info

Product

Resources

About